Representing large datasets typically requires significant abstraction when visualizing the representation on a computer display. Often, representing large datasets involves averaging, summing, or otherwise summarizing the data in the dataset to succinctly describe it in a manner appropriate for display upon a computer display. Such representations are often data lossy. Furthermore, known techniques that enable selection of portions of the representation to expand upon the details of the selected portion often provide the details in a list or text-based format that does not represent the context of the visualization. The known techniques do not progressively disclose a more granular level of detail without fundamentally shifting the visualization or introducing new panels. Furthermore, known techniques do not sufficiently organize data from multiple data sources into single visualizations.